Johns Hopkins engineer Suchi Saria has developed an AI-powered platform that is reducing sepsis mortality rates by 18% in dozens of hospitals across the United States—a significant advance in addressing a deadly immune response that claims about 270,000 lives each year.
Saria's Targeted Real-Time Early Warning System, or TREWS, integrates electronic health records with machine learning to help doctors diagnose sepsis cases nearly two hours earlier than traditional methods. Sepsis is easily missed because many of its symptoms, such as fever and confusion, are common in other medical conditions, she says. And in severe sepsis cases, an hour's delay in identification can be the difference between life and death.
As described in Nature Medicine, the system's earlier identification improves patient outcomes, reducing average hospital stays by half a day and intensive care unit use by 10% since its deployment in 2023.
Saria, an associate professor of computer science, biostatistics, health system informatics, and health policy and management, says that innovation like TREWS doesn't happen in isolation—it requires intentional investment and visionary support. She credits the National Science Foundation's Future of Work at the Human-Technology Frontier program with enabling her team to lay the groundwork for technologies that meaningfully augment clinical capacity by advancing methods that are scientifically rigorous and designed to integrate seamlessly into real-world care.
"At a time when clinicians are stretched thinner than ever, patients are older and sicker, and resources are finite, funding like this is essential. It powers the kind of breakthroughs that ensure health care evolves to meet tomorrow's demands without compromising the quality and humanity at its core," says Saria, the founding research director of the Whiting School of Engineering's Malone Center for Engineering in Healthcare and a member of the Johns Hopkins Data Science and AI Institute.
Saria—who has been recognized for her work on AI in health care by media outlets like TIME, Business Insider, Modern Healthcare, and Popular Science—began fighting the devastating effects of sepsis after losing her nephew to the disease in 2017.
TREWS has achieved a 90% adoption rate by seamlessly integrating into clinicians' workflows, leveraging comprehensive medical record data—including structured metrics like labs and vitals and unstructured insights from doctors' notes—to identify patients at risk of life-threatening complications and recommend treatment protocols. Amid national clinician shortages and an aging, increasingly complex patient population, TREWS streamlines key clinical workflows from emergency department admission through discharge, ensuring continuity of care during staff changes and department transfers.
Saria spun out this research through an NSF Small Business Innovation Research grant into a startup called Bayesian Health, transforming TREWS from an academic innovation into a scalable care augmentation platform capable of tackling many critical conditions beyond sepsis. Bayesian is now partnering directly with academic, regional, and community health systems, dramatically expanding the platform's reach by more than 800% last year.
"Federal funding, including vital support from the NSF, was foundational to the development of TREWS," Saria says. "It allowed us to conduct the kind of long-horizon, high-impact research that isn't immediately profitable but is essential for building technologies that truly improve patient outcomes. Without this funding, innovations like TREWS—born out of rigorous science and clinical collaboration—would not reach the bedside."
The consequences of losing this support are profound, she says.
"Institutions like Johns Hopkins, long known for pushing the boundaries of patient care, risk seeing their ability to innovate curtailed. More broadly, we risk creating an innovation gap where health care technology becomes stagnant, unable to keep pace with growing clinician shortages, rising patient complexity, and system-wide resource constraints. Continued investment is not optional—it's what ensures our health care systems can meet the future."
